1. Field of the Invention
The present invention relates to methods of use of facially amphiphilic polymers and oligomers, including pharmaceutical uses of the polymers and oligomers as antimicrobial agents and antidotes for hemorrhagic complications associated with heparin therapy. The present invention also relates to novel facially amphiphilic polymers and oligomers and their compositions, including pharmaceutical compositions. The present invention further relates to the design and synthesis of facially amphiphilic polymers and oligomers.
2. Related Art
Amphiphilic molecules exhibit distinct regions of polar and nonpolar character. These regions can result from substitution of hydrophobic and hydrophilic substituents into specific and distinct regions of conformationally defined molecules. Alternatively, a conformationally flexible molecule or macromolecule can adopt an ordered structure in which the hydrophobic and hydrophilic substituents on the molecule segregate to different areas or faces of the molecule. Commonly occurring amphiphilic molecules include surfactants, soaps, detergents, peptides, proteins and copolymers. These molecules have the capacity to self-assemble in appropriate solvents or at interfaces to form a variety of amphiphilic structures. The size and shape of these structures varies with the specific composition of the amphiphilic molecule and solvent conditions such as pH, ionic strength and temperature.
Amphiphilic peptides with unique broad-spectrum antimicrobial properties have been isolated from a variety of natural sources including plants, frogs, moths, silk worms, pigs and humans (H. G. Boman Immunol Rev. 2000 173:5-16; R. E. Hancock and R. Lehrer, Trends Biotechnol. 1998 16:82-88). These compounds include the magainin 1 and dermaseptin S1 isolated from the skin of frogs and the cecropin A isolated from the cecropia moth. These naturally occurring compounds have broad-spectrum antibacterial activity and they do not appear prone to the development of bacteria resistance. These compounds are of relatively low molecular weight and have a propensity to adopt an α-helical conformation in hydrophobic media or near a hydrophobic surface and as a result are facially amphiphilic, with one-third to two-thirds of the cylinder generated by the helical peptide has hydrophobic side chains while the remainder has hydrophilic side chains. The hydrophilic side chains are primarily positively-charged at neutral pH.
Hydrophobic amino acids compose 40-60% of the total number of residues in most anti-microbial peptides. The selectivity of the amphiphilic peptides (e.g. for bacteria vs. human erythrocytes) depends on the overall hydrophobicity. The biological activity of the compounds depends on the ratio of charged (c) to hydrophobic (h) residues. When the ratio is varied from 1:1 (c:h) to 1:2 (c:h) peptides with more hydrophobic residues tend to be more active toward erythrocyte membranes. Related peptides have been isolated from mammals and these anti-microbial peptides have been suggested to be an important component of the innate immune response. (Gennaro, R. et al. Biopoylmers (Peptide Science) 2000, 55, 31)
Secondary structures other than helices may also give rise to amphiphilic compounds. The protegrins are a related series of anti-microbial peptides. (J. Chen et al., Biopolymers (Peptide Science), 2000 55 88) The presence of a pair of disulfide bonds between Cys6-Cys15 and Cys8-Cys13 results in a monomeric amphiphilic anti-parallel β-sheet formed by the chain termini and linked by a β-turn. The amphiphilic β-sheet conformation is essential for anti-microbial activity against both gram-positive and gram-negative bacteria.
Following the initial discovery of cecropins and magainins, antimicrobial peptides have become a large and growing class of biologically interesting compounds (Zasloff, M., Curr. Opin. Immunol. 4:3-7 (1992); Zasloff, M., Trends Pharmacol. Sci. 21:236-238 (2000)). These compounds represent the first line of defense against microbes for many species, including plants, insects, worms, and mammals (Boman, H. G., Immunol. Rev. 173:5-16 (2000); Hancock, R. E., and Lehrer, R., Trends Biotechnol. 16:82-88 (1998)). In mammals, the peptides are produced and secreted by skin, mucosal surfaces and neutrophils. There are many different classes of natural host defense peptides (Zasloff, M., Curr. Opin. Immunol 4:3-7 (1992); Zasloff, M., Trends Pharmacol. Sci. 21:236-238 (2000); Steiner, H., et al., Nature, 292:246-248 (1981); Ganz, T., et al., Eur. J. Haematol. 44:1-8 (1990); Tang, Y. Q., et al., Science 286:498-502 (1999); Ganz, T., et al., J. Clin. Invest. 76:1427-1435 (1985); Landon, C., et al., Protein Sci. 6:1878-1884 (1997); Zhao, C., et al., FEBS Lett. 346:285-288 (1994); Peggion, E., et al., Biopolymers (Peptide Science) 43:419-431 (1998); Dempsey, C. E., Biochim. Biophys. Acta 1031:143-161 (1990)), but, in general, most contain between 20-40 amino acid residues and adopt an amphiphilic secondary structure as shown in FIG. 2.
The cytotoxic activity of the cationic and amphiphilic host defense peptides is also specific for bacteria over mammalian cells. This specificity is most likely related to fundamental differences between the two membrane types. For example, bacteria have a large proportion of negatively charged phospholipid headgroups on their surface, while, in contrast, the outer leaflet of animal cells is composed mainly of neutral lipids (Zasloff, M., Nature 415:389-395 (2002)). The presence of cholesterol in the animal cell membrane also appears to reduce the activity of the antimicrobial peptides.
The bactericidal activity of the host defense peptides is very rapid, occurring within minutes after exposure of bacteria to lethal doses of peptide. Several mechanisms have been proposed for the process of cell killing. According to the carpet mechanism, host defense peptides aggregate parallel to the membrane surface (Gazit, E., et al., Biohemistry 34:11479-11488 (1995); Pouny, Y., et al., Biochemistry 31:12416-12423 (1992)), leading to thinning and, ultimately, rupture of the membrane. In the so-called barrel-stave mechanism, the bound peptides on the cell surface self-associate into transmembrane helical bundles that form stable aqueous pores in the membrane (Merrifield, R. B., et al., Ciba Found. Symp. 186:5-20 (1994)). According to a third possible mechanism (DeGrado, W. F., et al., Biophys. J. 37:329-338 (1982)), the peptides initially bind only to the outer leaflet of the bilayer, leading to an increase in the lateral surface pressure of the outer leaflet relative to the inner leaflet of the bilayer. This pressure imbalance results in translocation of the peptides into the interior of the bilayer, with concomitant formation of transient openings in the membrane that allow hydration of the polar sidechains of the peptide and leakage of cellular contents. Most antimicrobial peptides probably act by more than one of these mechanisms. Additionally, some classes may interact with periplasmic or intercellular targets (Zasloff, M., Trends Pharmacol. Sci. 21:236-238 (2000)).
In addition to antibacterial activity, several of the host defense peptides possess antifungal activity. Examples of mammalian, insect and amphibian peptides with demonstrated antifungal activities include defensins, protegrins, lactoferrin-B, cecropins, and dermaseptins (DeLucca, A. J., and Walsh, T. J., Antimicob. Agents Chemother. 43:1-11 (1999)). The mechanism of cytotoxic action appears to be similar to that for bacteria, leading to rapid lysis of the fungal membrane.
Several host defense peptides also possess antiviral activity and inhibit the replication of both DNA and RNA viruses. See, for example, Sinha, S., et al., Antimicrob. Agents Chemother. 47:494-500 (2003); Belaid, A., et al., J. Med. Virol. 66:229-234 (2002); Egal, M., et al., Int. J. Antimicrob. Agents 13:57-60 (1999); Andersen, J. H., et al., Antiviral Rs. 51:141-149 (2001); and Bastian, A., and Schafer, H., Regul Pept. 15:157-161 (2001).
The human alpha-defensins have also been shown to inhibit the replication of HIV-1 isolates in vitro (Zhang, L., et al., Science 298: 995-1000 (2002). The antimicrobial peptides melittin and cecropin have also been reported to inhibit HIV-1 replication and it is suggested that they exert their activity by suppressing HIV gene expression (Wachinger, M., et al., J. Gen. Virol. 79:731-740 (1998)).
Although host defense peptides are found in a wide variety of species and are composed of many different sequences, their physiochemical properties are remarkably similar. They adopt an amphiphilic architecture with positively charged groups segregated to one side of the secondary structure and hydrophobic groups on the opposite surface. For example, magainin and some of the other naturally occurring antibacterial peptides contain positively charged amino acids and a large hydrophobic moment. Although these peptides exhibit considerable variation in their chain length, hydrophobicity and distribution of charges, they have a high propensity to adopt α-helical conformations in a hydrophobic environment, e.g., a cell surface or a natural or synthetic membrane (Oren, Z., and Shai, Y., Biopolymers (Peptide Science) 47:451-463 (1998)). The periodic distribution of hydrophobic and hydrophilic side chains in their amino acid sequences allows the segregation of the hydrophobic and hydrophilic side chains to opposite faces of the cylinder formed by the helix. These structures can be described as facially amphiphilic regardless of whether the secondary structure is a helix or sheet type fold. In fact, it is the overall physiochemical properties that are responsible for the biological activity of these peptides and not the precise sequence (Zasloff, M., Curr. Opin. Immunol. 4:3-7 (1992); Zasloff, M., Trends Pharmacol. Sci. 21:236-238 (2000); Hancock, R. E., and Lehrer, R., Trends Biotechnol. 16:82-88 (1998); DeGrado, W. F., et al., J. Amer. Chem. Soc. 103:679-681 (1981); DeGrado, W. F., Adv. Prot. Chem. 39:51-124 (1988); Tossi, A., et al., Biopolymers 55:4-30 (2000); Merrifield, E. L., et al., Int. J. Pept. Protein Res. 46:214-220 (1995); Merrifield, R. B., et al., Proc Natl Acad Sci (USA) 92:3449-3453 (1995)). Thus, facial amphiphilicity, i.e., the alignment of polar (hydrophilic) and nonpolar (hydrophobic) side chains on opposite faces of a secondary structural element formed by the peptide backbone, and not amino acid sequence or any particular secondary/tertiary structure, chirality or receptor specificity, is responsible for the biological activity of these peptides.
The design of non-biological polymers with well-defined secondary and tertiary structures has received considerable attention in the past few years (Gellman, S. H., Acc. Chem. Res. 31:173-180 (1998); Barron, A. E., and Zuckermann, R. N., Curr. Opin. Chem. Biol. 3:681-687 (1999); Stigers, K. D., et al., Curr. Opin. Chem. Biol., 3:714-723 (1999)). Using these principles, investigators have designed synthetic antimicrobial peptides by idealizing the amphiphilic α-helical arrangement of sidechains observed in the natural host defense peptides, leading to a large number of potent and selective antimicrobial compounds (Tossi, A., et al., Biopolymers 55:4-30 (2000); DeGrado, W. F., Adv. Protein. Chem. 39:51-124 (1988); Maloy, W. L., and Kari, U. P., Biopolymers 37:105-122 (1995); Zasloff, M., Curr. Opin. Immunol. 4:3-7 (1992); Boman, H. G., et al., Eur. J. Biochem. 201:23-31 (1991); Oren, Z., and Shai, Y., Biopolymers 47:451-463 (1998)).
β-peptides have also provided another avenue to test and further elucidate the features required for the construction of bactericidal agents. β-peptides adopt L+2 helices, which have an approximate 3-residue geometric repeat. Thus, if polar and apolar sidechains are arranged with precise three-residue periodicity in the sequence of a β-peptide, they should segregate to opposite sides of the helix. Using this approach, DeGrado and co-workers (Hamuro, Y., et al., J. Amer. Chem. Soc. 121:12200-12201 (1999); Liu, D., and DeGrado, W. F., J. Amer. Chem. Soc., 123:7553-7559 (2001)) have designed synthetic β-peptide oligomers that are roughly equipotent in antimicrobial activity to many naturally occurring peptide antibiotics. The antimicrobial activities of these β-peptides and their specificities for bacterial cells over mammalian cells can be controlled by fine-tuning their hydrophobicities and chain lengths. Gellman and coworkers have also synthesized cyclically constrained β-peptides possessing potent antimicrobial activity and minimal activity against mammalian cells (Porter, E. A., et al., Nature 404:565 (2000)).
Non-peptidic antimicrobial polymers have also been developed. For example, suitably substituted polymers lacking polyamide linkages that are capable of adopting amphiphilic conformations have been designed and synthesized. Solid phase chemistry technology has been utilized to synthesize a class of meta substituted phenylacetylenes that fold into helical structures in appropriate solvents (Nelson, J. C., et al, Science 277:1793-1796 (1997); Prince, R. B., et al., Angew. Chem. Int. Ed. 39:228-231 (2000)). These molecules contain an all hydrocarbon backbone with ethylene oxide side chains such that when exposed to a polar solvent (acetonitrile), the backbone collapses to minimize its contact with this polar solvent. As a result of the meta substitution, the preferred folded conformation is helical. This helical folding is attributed to a “solvophobic” energy term; although, the importance of favorable π-π aromatic interactions in the folded state are also likely to be important. Furthermore, addition of a less polar solvent (CHCl3) results in an unfolding of the helical structure demonstrating that this folding is reversible.
Regioregular polythiophenes (5 and 6) have been shown to adopt amphiphilic conformations in highly ordered π-stacked arrays with hydrophobic side chains on one side of the array and hydrophilic side chains on the other side. These polymers form thin films useful in the construction of nanocircuits. (Bj∅rnholm et al., J. Am. Chem. Soc., 1998 120, 7643) These materials would be facially amphiphilic as defined herein; however, no biological properties have been reported for these compounds.

Antimicrobial peptides have been incorporated onto surfaces or bulk materials, with some retention of antimicrobial properties. Haynie and co-workers at DuPont have investigated the activity of antibacterial peptides that have been covalently attached to solid surfaces (S. L. Haynie et al., Antimicrobial Agents Chemother., 1995 39:301-7; S. Margel et al., J. Biomed. Mater. Res., 1993, 27:1463-76). A variety of natural and de novo designed peptides were synthesized and tested for activity while still attached to the solid support. The activity of the peptides decreased when attached to the solid support although the peptides retained their broad spectrum of activity. For example, a de novo designed peptide referred to as E14LKK was reported to have a MBC (minimum bactericidal activity) of 31 μg/ml in solution as opposed to 1.5 mg/ml when attached to a solid phase bead. The peptides were attached to the resin with a 2 to 6-carbon alkyl linker. The porosity of Pepsyn K, the resin used in the synthesis, is small (0.1 to 0.2 μm) compared to the bacteria, so the microbes may have been unable to penetrate into the interior of the resin. Thus the great majority of the peptide would not be available for binding to cells. The antimicrobial activity did not arise from a soluble component; no leached or hydrolyzed peptide was observed and the soluble extracts were inactive. These studies indicate quite convincingly that antimicrobial peptides retain their activity even when attached to a solid support. However, there is a need to optimize the presentation of the peptides to increase their potency.
Other antimicrobial polymeric materials have been reported which contain chemical functionalities known to be antimicrobial (J. C. Tiller et al., Proc Natl Acad Sci USA, 2001 98:5981-85). A large portion of this work uses chemical functions such as alkylated pyridinium derivatives, which are known to be toxic to mammalian cells. The antibiotic ciprofloxacin has been grafted into a degradable polymer backbone (G. L. Y. Woo, et al., Biomaterials 2000 21:1235-1246). The activity of this material relies on cleavage of the active component from the polymer backbone.
In addition, Mandeville et al., U.S. Pat. No. 6,034,129, disclose anti-infective vinyl copolymers, wherein monomers with hydrophobic and hydrophilic side chains have been randomly polymerized to produce polymers with amphiphilic properties. These materials are produced by polymerization of hydrophobic and hydrophilic acrylate monomers. Alternately, the hydrophobic side chain is derived from a styrene derivative which is copolymerized with a hydrophilic acrylate monomer wherein an ionic group is linked to the carboxylic, acid.
Tew et al. (Tew, G. N., et al., Proc. Natl. Acad. Sci. (USA) 99:5110-5114 (2002)) disclose the design and synthesis of a series of biomimetic, facially amphiphilic arylamide polymers possessing antimicrobial activity. The arylamide polymers were designed using de novo computational design techniques.
WIPO Publ. No. WO 02/100295 discloses facially amphiphilic polyamide, polyester, polyurea, polycarbonate, and polyurethane polymers with anti-infective activity, and articles made from them having biocidal surfaces. WIPO Publ. No. WO 02/100295 is fully incorporated by reference herein in its entirety.
WIPO Publ. No. WO 02/072007 discloses a number of facially amphiphilic polyphenylene and heteroarylene polymers, including polyphenylalkynyl polymers, with anti-infective activity and articles made therefrom having biocidal surfaces. WIPO publication no. WO 02/072007 is fully incorporated by reference herein in its entirety.
An alternative method to make amphiphilic polymers is to produce block copolymers comprised of hydrophobic blocks (A) and hydrophilic blocks (B), commonly polypropyleneoxy and polyethylenoxy segments respectively, into A-B, A-B-A or similar copolymers. These copolymers also are not facially amphiphilic as defined herein.